PROBABILITY, STATISTICS AND LINEAR PROGRAMMING LAB

PAPER CODE: BS-252

Lab Manual
 INSTITUTE VISION
To develop a leaning environment for academic excellence with technological and management
competence at par with international standards and to nurture young minds with high spiritual and
ethical values.

INSTITUTE MISSION
The Institute will focus to incorporate following agendas in the teaching methodology:

Engineering Hardware – Software Symbiosis

Practical exercises in all Engineering and Management disciplines will be carried out using
Hardware equipment as well as the related software, enabling deeper understanding of basic
concepts and encouraging inquisitive nature.

Life – Long Learning

The Institute strives to match technological advancements and encourage students to keep updating
their knowledge for enhancing their skills and inculcating their habit of continuous learning.
Liberalization and Globalization
The Institute endeavors to enhance technical and management skills of students so that they are
intellectually capable and competent professionals with industrial aptitude to face the challenges of
globalization.
Diversification
The Engineering, Technology and Management disciplines have diverse fields of studies with
different attributes. The aim is to create a synergy of the above attributes by encouraging analytical
thinking.
Digitization of Learning Processes
The Institute provides seamless opportunities for innovative learning in all Engineering and
Management disciplines through digitization of learning processes using analysis, synthesis,
simulation, graphics, tutorials and related tools to create a platform for multi-disciplinary approach.
Entrepreneurship
The Institute strives to develop potential Engineers and Managers by enhancing their skills and
research capabilities so that they become successful entrepreneurs and responsible citizens.

PROBABILITY, STATISTICS AND LINEAR PROGRAMMING LAB

PAPER CODE: BS-252

COURSE OBJECTIVE

1. To understand probability and probability distributions.

2. To understand methods of summarization and visualization of data.

3. To understand the use of various hypothesis-testing techniques.

4. To understand methods for solving linear programming problems.

COURSE OUTCOMES

At the end of the course student will be able to:

C.252.1 Apply probability concepts and describe probability distributions to solve

real-world problems. (Unit I)

C.252.2 Analyze joint distributions and visualize data using statistical tools.

(Unit II)

C.252.3 Evaluate hypothesis-testing techniques and apply regression and

correlation methods for analyzing data and making predictions. (Unit III)

C.252.4 Formulate and solve linear programming problems using mathematical

methods. (Unit IV)

No Title of Lab Experiments CO
.
1. Demonstrations of SCILAB to simple programming concepts like CO1
addition, multiplication and transpose of matrices (using loops)
2. CO1
Fitting of Binomial distribution for given n and p

3. Fitting of Poisson distribution after computing mean CO1

4. Fitting Standard Normal Distribution CO1

5. CO2
Compute covariance and correlation for two datasets

6. CO2
Write a program to visualize statistical data using histogram.

7. CO3
Fitting of regression lines

8. CO3
Perform a Chi-square test for goodness of fit for a given dataset

9. CO3
Perform t-test for comparing means of two samples.

10. Program for forming first simplex table for a three variable linear CO4
programming problem.
11. CO4
Program for solving Transportation Problem of three variables.

12. CO4
Program for solving Assignment Problem of three variables.

EVALUATION SCHEME
Laboratory
Components Internal External
Marks 40 60
Total Marks 100
GUIDELINES FOR CONTINUOUS ASSESSMENT FOR EACH EXPERIMENT

Attendance and performance in experiments – 40 marks

 Practical performance
 Result validation
 Attendance and Viva Questions
 Timely Submission of Lab records

The Rubrics for Experiment execution, Lab file and viva voice is given below:

Experiment Marks details:

Experiment Grading Rubrics

LAC/ Max.
Component
Sr No. Marks
(LAC) 2 marks 1 mark

Develops the program and

Practical Not able to complete the
LAC 1 2 exhibits proficiency in
Performance program in the required way.
execution.

Accuracy of results.
Results with moderate
Demonstrates excellent
Result accuracy. Demonstrates partial
LAC 2 2 understanding of the
Validation understanding of the concepts
concepts relevant to the
relevant to the program.
program.
One mark for attendance.

Attendance Three marks for answering more than 75% questions.

LAC 3 and Viva 4
Questions
Two marks for answering more than 50% questions.

One mark for answering less than 50% questions.

Timely
Submission
LAC 4 2 On time submission Late submission
of Lab
Records
CONTENTS
Experiment 1(a): Matrix Addition ........................... ………....7
Experiment 1(b): Matrix Multiplication.................................. 9
Experiment 1(c): Matrix Transpose ...................................... 11
Experiment 2: Fitting Binomial distribution for given N and P…..13
Experiment 3: Fitting Poisson distribution By computing mean….16
Experiment 4: Plotting Standard Normal distribution Curve…..19
Experiment 5: Compute Covariance and Correlation for 2 datasets…..23
Experiment 6: Visualize Statistical Data Using Histogram ... 25
Experiment 7: Fitting Regression Lines ................................ 27
Experiment 8: Perform a chi-square test for goodness to fit for a given data….31
Experiment 9: Perform t-test for comparing means of two samples….34
Experiment 10: Solving a LPP using Simplex Method ......... 38
Experiment 11: Solving Transportation Problem .................. 44
Experiment 12: Solving a 3 by 3Assignment Problem .......... 59
EXPERIMENT 1(A): MATRIX ADDITION

Aim/Objective: Program for finding matrix addition

Course Outcome: CO1
Software Used: SCILAB
Theory: A Matrix is a rectangular array of numbers, expressions or symbols, arranged in
rows and columns. Matrix addition is commutative and associative just like addition of
arithmetic numbers. Addition or subtraction can be performed on two matrices if and only if
they are of same order.
𝑎11 𝑎12 𝑎13 𝑏11 𝑏12 𝑏13
𝐴 = (𝑎21 𝑎22 𝑎23 ) 𝐵 = (𝑏21 𝑏22 𝑏23 )
𝑎31 𝑎32 𝑎33 𝑏31 𝑏32 𝑏33
𝑎11 ± 𝑏11 𝑎12 ± 𝑏12 𝑎13 ± 𝑏13
Then 𝐴 ± 𝐵 = (𝑎21 ± 𝑏21 𝑎22 ± 𝑏22 𝑎23 ± 𝑏23 )
𝑎31 ± 𝑏31 𝑎32 ± 𝑏32 𝑎33 ± 𝑏33
Matrix Addition Algorithm:
 Declare variables and initialize necessary variables
 Enter the elements of matrices row wise using loops
 Add the corresponding elements of the matrices using nested loops
 Print the resultant matrix as console output
5 1
1 3 0 2
2 2
Sample problem: Add the matrices 𝐴 = ( 5 −1 3) , 𝐵 = ( 2
−1
0 1)
2
3 3 3 −2 −1 0
1 3 5
Solution: 𝐶 = 𝐴 + 𝐵 = ( 2 −1 4)
1 2 3
CODE:
m=input("enter number of rows of the Matrix: ");
n=input("enter number of columns of the Matrix: ");
disp('enter the first Matrix')
fori=1:m
forj=1:n
A(i,j)=input('\');
end
end
disp('enter the second Matrix')
fori=1:m

[7]
forj=1:n
B(i,j)=input('\');
end
end
fori=1:m
forj=1:n
C(i,j)=A(i,j)+B(i,j);
end
end
disp('The first matrix is')
disp(A)
disp('The Second matrix is')
disp(B)
disp('The sum of the two matrices is')

disp(C)

[8]
EXPERIMENT 1(B): MATRIX MULTIPLICATION

Aim/Objective: Program to compute matrix multiplication

Course Outcome: CO1
Software Used: SCILAB
Theory: Matrix product 𝐴𝐵 is possible only if number of columns in matrix 𝐴 are same as
number of rows in matrix 𝐵.
𝑏11 𝑏12
𝑎11 𝑎12 𝑎13
𝐴 = (𝑎 𝑎 𝑎 ) 𝐵 = (𝑏21 𝑏22 )
21 22 23
𝑏31 𝑏32
𝑐 = 𝑎11 𝑏11 + 𝑎12 𝑏21 + 𝑎13 𝑏31 𝑐12 = 𝑎11 𝑏12 + 𝑎12 𝑏22 + 𝑎13 𝑏32
𝐶 = 𝐴𝐵 = ( 11 )
𝑐21 = 𝑎21 𝑏11 + 𝑎22 𝑏21 + 𝑎23 𝑏31 𝑐22 = 𝑎21 𝑏12 + 𝑎22 𝑏22 + 𝑎23 𝑏32
Note that: (i) 𝐴𝑚×𝑛 𝐵𝑛×𝑘 = 𝐶𝑚×𝑘
(ii) 𝐴𝐵 ≠ 𝐵𝐴 in general
Matrix Multiplication Algorithm:
 Declare variables and initialize necessary variables
 Enter the elements of matrices row wise using loops
 Check the number of rows and column of first and second matrices
 If number of rows of first matrix is equal to the number of columns of second
matrix, go to next step. Otherwise, print ’Matrices are not conformable for
multiplication’ and abort.
 Multiply the matrices using nested loops
 Print the product in matrix form as console output
sin 𝑥 cos 𝑥 sin 𝑥 sin 𝑥
Sample Problem: If 𝐴 = ( )𝐵 = ( )
sin 𝑥 cos 𝑥 cos 𝑥 cos 𝑥
Find 𝐴𝐵and 𝐵𝐴
2 2 2 2 1 1
Solution:𝐴𝐵 = (sin2 𝑥 + cos 2 𝑥 sin2 𝑥 + cos 2 𝑥 ) = ( )
sin 𝑥 + cos 𝑥 sin 𝑥 + cos 𝑥 1 1
2
𝐵𝐴 = (2 sin 𝑥 sin 2𝑥 )
sin 2𝑥 2 cos 2 𝑥
CODE:
m=input("Enter number of rows of the first Matrix: ");
n=input("Enter number of columns of the first Matrix: ");
p=input("Enter number of rows of the second Matrix: ");
q=input("Enter number of columns of the second Matrix: ");

[9]
ifn==p
disp('Matrices are conformable for multiplication')
else
disp('Matrices are not conformable for multiplication')
break;
end
disp('enter the first Matrix')
fori=1:m
forj=1:n
A(i,j)=input('\');
end
end
disp('enter the second Matrix')
fori=1:p
forj=1:q
B(i,j)=input('\');
end
end
C=zeros(m,q);
fori=1:m
forj=1:q
fork=1:n
C(i,j)=C(i,j)+A(i,k)*B(k,j);
end
end
end
disp('The first matrix is')
disp(A)
disp('The Second matrix is')
disp(B)
disp('The product of the two matrices is')
disp(C)

[10]
EXPERIMENT 1(C): MATRIX TRANSPOSE

Aim/Objective: Program to compute matrix transpose

Course Outcome: CO1
Software Used: SCILAB
Theory: The matrix 𝐴′ or 𝐴𝑇 obtained by interchanging rows and columns of a matrix 𝐴 is
known as its transpose.
1 3 5 1 2 0
𝐴 = ( 2 −1 4) 𝐴𝑇 = ( 3 −1 2)
0 2 3 5 4 3
Matrix Transpose Algorithm:
 Declare variables and initialize necessary variables
 Enter the elements of matrix by row wise using loop
 Interchange rows to columns using nested loops
 Print the transposed matrix as console output
CODE:
clc
m=input("Enter number of rows of the Matrix: ");
n=input("Enter number of columns of the Matrix: ");
disp('Enter the Matrix')
for i=1:m
[11]
for j=1:n
A(i,j)=input('\');
end
end
B=zeros(n,m);
for i=1:n
for j=1:m
B(i,j)=A(j,i)
end
end
disp('Entered matrix is')
disp(A)
disp('Transposed matrix is')
disp(B)

[12]
EXPERIMENT 2: FITTING BINOMIAL DISTRIBUTION FOR GIVEN N AND P

Aim/Objective: Program for fitting Binomial distribution when n and p are given
Course Outcome: CO1
Software Used: SCILAB
Theory: A series of independent trials which result in one of the two mutually exclusive
outcomes ‘success’ or ‘failure’ such that the probability of the success (or failure) in each
trials is constant, then such repeated independent trials are called as ‘Bernoulli trials’. A
discrete random variable which results in only one of the two possible outcomes (success or
failure) is called Binomial variable.
Let there be 𝑛 independent finite trials in an experiment such that
i. Each trial has only two possible outcomes success and failure
ii. Probability of success (𝑝) and probability of failure (𝑞) are constant for all the trials
and 𝑝 + 𝑞 = 1.
Then if a random variable 𝑋 denotes the number of successes in 𝑛 trials, then
𝑛
𝑃 (𝑋 = 𝑟 ) = 𝐶𝑟 𝑝𝑟 𝑞 𝑛−𝑟 or 𝑃(𝑟) = 𝑛
𝐶𝑟 𝑞 𝑛−𝑟 𝑝𝑟
∴ Binomial distribution may be given as (𝑞 + 𝑝)𝑟
Sample Problem: Fit a binomial distribution to the following data and compare theoretical
frequencies with actual ones
𝑥 0 1 2 3 4 5 6 7 8 9
𝑓 6 20 28 12 8 6 0 0 0 0
∑ 𝑓𝑥
Solution: Mean of the given distribution = ∑𝑓
, ∑ 𝑓 = 80
0 + 20 + 56 + 36 + 32 + 30 + 0 87
= = = 2.175
80 40
Let mean of binomial distribution to be fitted = 𝑛𝑝 = 2.175
Also 𝑛 = 10 ∴ 𝑝 = 0.2175 𝑞 = 1 − 0.2175 = 0.7825
∴ B.D. is given by 80(0.7825 + 0.2175 )10
Theoretical frequencies using binomial distribution are given in the table below:
𝑥 P(r) = 𝑛 𝐶𝑟 𝑞 𝑛−𝑟 𝑝𝑟 Theoretical frequencies (𝑓) = 80× 𝑃(𝑟)
10 (
0 𝐶0 0.7825)10 (0.2175)0 = 0.086 6.9 = 7(say)
10 (
1 𝐶1 0.7825)9 (0.2175)1 =0.239 19.1 = 19(say)
10 )8 ( )2
2 𝐶2 (0.7825 0.2175 = 0.299 23.9 = 24(say)
10
3 𝐶3 (0.7825)7 (0.2175)3 = 0.22 17.8 = 18(say)
10 )6 ( )4
4 𝐶4 (0.7825 0.2175 = 0.11 8.6 = 9(say)
10
5 𝐶5 (0.7825)5 (0.2175)5 = 0.04 2.9 = 3(say)

[13]
 10
6 𝐶6 (0.7825)4 (0.2175)6 = 0.008 0.66 = 0(say)
10
7 𝐶7 (0.7825)3 (0.2175)7 = 0.001 0.11 = 0(say)
10
8 𝐶8 (0.7825)2 (0.2175)8 = 0 0
10 (
9 𝐶9 0.7825)1 (0.2175)9 = 0 0

Algorithm for fitting Binomial distribution when n and p are given:

 Input n and p
 Enter the elements 𝑥𝑖 using loop
 Enter the corresponding frequencies 𝑓𝑖 using loop
 Calculate theoretical frequencies using binomial function
 Plot the graph between expected and theoretical frequencies using plot2d command
CODE:
Clearall
disp("enter no of observation")
n=input('\')
disp("value of p")
p=input('\')
disp("enter the vlaue of x")
fori=1:n
X(1,i)=input('\')
end
disp("enter no of frequency")
forj=1:n
F(1,j)=input('\')
end
EF=sum(F)*binomial(p,n-1)
disp("Given frequencies")
disp(F)
disp("Expected frequencies")
disp(EF)
plot2d3(0:n-1,F)
plot2d(0:n-1,EF)

[14]
[15]
EXPERIMENT 3: FITTING POISSON DISTRIBUTION BY COMPUTING MEAN

Aim/Objective: Program for fitting Poisson distribution after computing mean

Course Outcome: CO1
Software Used: SCILAB
𝑒 −𝜆 𝜆𝑟
Theory: Poisson distribution with𝑃(𝑟) = 𝑟! is a limiting case of Binomial distribution,
under the conditions 𝒊. 𝑛 → ∞ 𝒊𝒊. 𝑝 → 0𝒊𝒊𝒊. 𝑛𝑝 = 𝜆 is finite
Sample Problem: A skilled typist kept a record of his mistakes made per day during 300
working days. Fit a Poisson distribution to compare theoretical frequencies with actual ones.
Mistakes per day 0 1 2 3 4 5 6
Number of days 143 90 42 12 9 3 1
∑ 𝑓𝑥
Solution: Mean of the given distribution= ∑𝑓
, ∑ 𝑓 = 300
0 + 90 + 84 + 36 + 36 + 15 + 6 89
= = = 0.89 = 𝜆
300 100
𝑒 −𝜆 𝜆𝑟 Theoretical frequency
Mistakes per day 𝑃 (𝑟 ) =
𝑟! 300× 𝑃(𝑟)
−(0.89) 0
𝑒 (0.89)
0 = 0.411 123.3=123 (say)
0!
𝑒 −(0.89) (0.89)1
1 = 0.365 109.5=110 (say)
1!
𝑒 −(0.89) (0.89)2
2 = 0.163 48.9=49 (say)
2!
𝑒 −(0.89) (0.89)3
3 = 0.048 14.4=14 (say)
3!
𝑒 −(0.89) (0.89)4
4 = 0.011 3.3=3 (say)
4!
𝑒 −(0.89) (0.89)5
5 = 0.002 0.6=1 (say)
5!
𝑒 −(0.89) (0.89)6
6 = 0.0003 0.09=0 (say)
6!

Algorithm for fitting Poisson distribution after computing mean:

 Input the number of observations
 Enter the elements 𝑥𝑖 using loop
 Enter the corresponding frequencies 𝑓𝑖 using loop
𝑒 −𝜆 𝜆𝑥𝑖
 Calculate theoretical frequencies using 𝑃(𝑥𝑖 ) =
𝑟!
 Display 𝑃(𝑥𝑖 )
[16]
  Plot the graph between expected and theoretical frequencies using plot2d command

CODE:

clc
clear
disp("enter no of observation")
n=input('\')
disp("enter the vlaue of x")
for i=1:n
X(1,i)=input('\')
end
disp("enter no of frequency")
for j=1:n
F(1,j)=input('\')
end
disp("Mean of the distribution is")
M=sum(F.*X)/sum(F)
disp(M)
for i=1:n
P(1,i)=sum(F)*exp(-M)*M^(X(i))/factorial(X(i))
end
disp("Expected frequencies are")
disp(P)
plot2d(X,P)

[17]
[18]
EXPERIMENT 4: PLOTTING STANDARD NORMAL DISTRIBUTION CURVE

Aim/Objective: Plotting Standard Normal curve for given parameter values

Course Outcome: CO1
Software Used: SCILAB
Theory: The normal distribution developed by Gauss is a continuous distribution and is very
useful in practical applications. It can be considered as the limiting form of the Binomial
Distribution when the number of trials (𝑛), is very large and neither 𝑝 nor𝑞is very small. The
probability curve of a normal variate𝑥 with mean𝜇 and standard deviation𝜎 is given by:
1 𝑥−𝜇 2
1 − ( )
𝑝 (𝑥 ) = 𝜎 𝑒 2 𝜎 , −∞ < 𝑥 < ∞
√2𝜋
Any normal variate 𝒙 with mean 𝝁 and
standard deviation 𝝈 is changed to a
𝑥−𝜇
standard normal variate 𝑧 = 𝜎 , and
hence the probability density function
of 𝑧 is given by:
1 2
1
𝜙 (𝑧 ) = 𝑒 − 2𝑧 , −∞ < 𝑧 < ∞
√ 2𝜋
The normal distribution with mean 𝜇 and
variance 𝜎 2 is denoted by 𝑁(𝜇 , 𝜎 2 ).
Adjoining figure shows a normal
distribution curve for standard normal
variate 𝑧.
Sample Problem: The daily wages of
1000 workers are normally distributed
with mean 100$ and standard deviation
5$. Estimate the number of workers whose daily wages will be: (𝑖) between 100$ and 105$
(𝑖𝑖) between 96$ and 105$
(𝑖𝑖𝑖) more than 110$ (𝑖𝑣) more than 110$
Solution: Let the random variable 𝑋 denote the daily wages of workers in dollars.
Then 𝑋 is a random variable with mean 𝜇 = 100 and S.D. 𝜎 = 5.
𝑋−𝜇 𝑋−100
i.e. 𝑋~𝑁(100 , 25) and 𝑧 = =
𝜎 5
⇒ 𝑋 = 100 + 5𝑧
(𝑖) 𝑃(100 < 𝑋 < 105)
= 𝑃(100 < 100 + 5𝑧 < 105)
100−100 105−100
= 𝑃( <𝑧< )
5 5
= 𝑃(0 < 𝑧 < 1) = 0.3413using Z table

[19]
(𝑖𝑖) 𝑃(96 < 𝑋 < 105)
= 𝑃(96 < 100 + 5𝑧 < 105)
96 − 100 105 − 100
= 𝑃( <𝑧< )
5 5
= 𝑃(−0.8 < 𝑧 < 1)
= 𝑃(0 < 𝑧 < 0.8) + 𝑃(0 < 𝑧 < 1)
= 0.2881 + 0.3413 = 0.6294 using Z table
(𝑖𝑖𝑖) 𝑃(𝑋 > 110)
= 𝑃(100 + 5𝑧 > 110)
110 − 100
= 𝑃 (𝑧 > )
5
= 𝑃 (𝑧 > 2)
= 0.5 − 𝑃(0 < 𝑧 < 2)
= 0.5 − 0.4772 = 0.0228 using Z table
(𝑖𝑣) 𝑃 (𝑋 < 92)
= 𝑃(100 + 5𝑧 < 92)
92 − 100
= 𝑃 (𝑧 < )
5
= 𝑃(𝑧 < −1.6)
= 𝑃(𝑧 > 1.6) (Symmetry)
= 0.5 − 𝑃(0 < 𝑧 < 1.6)
= 0.5 − 0.4452 = 0.0548 using Z table
Algorithm:

1. Input:

 A dataset of observations (sample data).

2. Calculate the sample mean and sample standard deviation:

 The sample mean 𝝁 is the average of all data points.

 The sample standard deviation 𝝈 is the square root of the variance, which is the
average of the squared differences from the mean.

3. Plot the histogram of the data:

 This helps visualize the distribution of the data.

4. Fit the normal distribution:

[20]
  Use the sample mean 𝝁 and sample standard deviation 𝝈 to fit a normal distribution.

5. Plot the fitted normal distribution curve:

 Overlay the normal distribution on the histogram of the data.

CODE:
// Clear workspace
clear;
clc;

// Input the dataset

x=[5.1,5.5,5.8,6.0,6.1,6.2,6.3,6.5,6.6,6.8];// Example data

// Number of data points

n=length(x);

// Step 1: Compute mean and standard deviation

mu=mean(x);
sigma=stdev(x);

// Step 2: Define the range for the normal distribution

xmin=min(x)-1;
xmax=max(x)+1;
x_fit=linspace(xmin,xmax,100);

// Step 3: Compute the fitted normal distribution

pdf_fit=(1/(sigma*sqrt(2*%pi)))*exp(-((x_fit-mu).^2)/(2*sigma^2));

// Step 4: Plot histogram and fitted normal distribution

clf;// Clear figure
histplot(10,x);// Histogram with 10 bins
plot(x_fit,pdf_fit*n*(xmax-xmin)/10,'r-','LineWidth',2);// Scale PDF to histogram
xlabel("Data");
ylabel("Frequency");
title("Fitting a Normal Distribution");
legend(["Histogram","Fitted Normal Distribution"],"location","upper_right");

[21]
[22]
EXPERIMENT 5: COMPUTE COVARIANCE AND CORRELATION FOR 2 DATASETS

Aim/Objective: Program for computing Covariance and Correlation

Course Outcome: CO2
Software Used: SCILAB
Theory: Correlation is a measure of association between two variables; which may be
dependent or independent. Whenever two variables 𝑥 and 𝑦 are so related; that increase in
one is accompanied by an increase or decrease in the other, then
the variables are said to be correlated. Coefficient of correlation
(𝑟) lies between −1 and +1, i.e.−1 ≤ 𝑟 ≤ 1.
If 𝑟 is zero; no correlation between two variables, positive
correlation (0 < 𝑟 ≤ +1); when both variables increase or
decrease simultaneously, and negative correlation (−1 ≤ 𝑟 < 0);
when increase in one is associated with decrease in other variable
and vice-versa.
4.4 Karl Pearson Coefficient of Correlation
Coefficient of correlation (𝑟) between two variables 𝑥 and 𝑦 is defined as
𝐶𝑜𝑣𝑎𝑟𝑖𝑎𝑛𝑐𝑒 (𝑥,𝑦) ∑ 𝑑𝑥 𝑑𝑦 𝜌
𝑟= = =𝜎
√𝑉𝑎𝑟𝑖𝑎𝑛𝑐𝑒 (𝑥) √𝑉𝑎𝑟𝑖𝑎𝑛𝑐𝑒 (𝑦) √(∑ 𝑑𝑥2 )(∑ 𝑑𝑦
2) 𝑥 𝜎𝑦

where 𝑑𝑥 = 𝑥 − 𝑥 , 𝑑𝑦 = 𝑦 − 𝑦 , 𝑥, 𝑦 are means of 𝑥 and 𝑦 data values.

∑ 𝑑𝑥 𝑑𝑦
𝜌 = 𝐶𝑜𝑣 (𝑥, 𝑦) = is the covariance between the variables 𝑥 and 𝑦.
𝑛
2
∑ 𝑑𝑥2 ∑ 𝑑𝑦
Also 𝜎𝑥 = √ and 𝜎𝑦 = √
𝑛 𝑛

Sample Problem: If 𝐶𝑜𝑣 (𝑥, 𝑦) = 10, 𝑣𝑎𝑟(𝑥 ) = 25, 𝑣𝑎𝑟(𝑦) = 9 find coefficient of
correlation.
𝐶𝑜𝑣 (𝑥,𝑦) 10 10
Solution: 𝑟 = = = 5×3 = 0.67
√𝑉𝑎𝑟(𝑥) √𝑉𝑎𝑟(𝑦) √25 √ 9

Algorithm:
1. Input two datasets as vectors.
2. Compute the mean of each dataset.
3. Calculate the covariance using the formula:
Cov(X, Y) = sum((X - mean(X)) .* (Y - mean(Y))) / (n - 1)
4. Compute the correlation using the formula:
Corr(X, Y) = Cov(X, Y) / (std(X) * std(Y))
5. Display the results.
Code:
// Define the datasets

[23]
X=[1,2,3,4,5];// Example dataset 1
Y=[2,4,6,8,10];// Example dataset 2

// Number of data points

n=length(X);

// Compute the means

mean_X=mean(X);
mean_Y=mean(Y);

// Compute the covariance

covariance=sum((X-mean_X).*(Y-mean_Y))/(n-1);

// Compute the standard deviations

std_X=stdev(X);
std_Y=stdev(Y);

// Compute the correlation

correlation=covariance/(std_X*std_Y);

// Display the results

disp("Covariance:");
disp(covariance);
disp("Correlation:");

disp(correlation);

[24]
EXPERIMENT 6: VISUALIZE STATISTICAL DATA USING HISTOGRAM

Aim/Objective: Program to visualize statistical data using histogram:

Course Outcome: CO2
Software Used: SCILAB
Theory: A histogram is a visual representation of the distribution of quantitative data. It is
represented by a set of rectangles, adjacent to each other, where each bar represents a kind of
data.

Algorithm:
1. Input the dataset to be visualized.
2. Define the number of bins for the histogram.
3. Use the histplot function to generate and display the histogram.
4. Customize the plot with labels and title for better visualization.

Code:
// Define the dataset
data=[5,7,8,9,6,7,9,8,10,7,6,8,9,5,6,7,8,9,10,7];

// Define the number of bins

num_bins=5;// Adjust this to change the bin granularity

// Generate and display the histogram

clf();// Clear any existing plots
histplot(num_bins,data);

title("Histogram of Dataset");
xlabel("Data Values");
ylabel("Frequency");

// Optionally, save the plot

// xs2png(gcf(), "histogram.png");

[25]
[26]
EXPERIMENT 7: FITTING REGRESSION LINES

Aim/Objective: Program for Regression lines

Course Outcome: CO3
Software Used: SCILAB
Theory: Regression describes the functional relationship between dependent and
independent variables; which helps us to make estimates of one variable from the other. If we
plot the observations of the linear regression between
two variables, actually two straight lines can
approximately be drawn through the scatter diagram.
One line estimates values of 𝑦 for specified values of
𝑥 (known as line of regression of 𝑦 on 𝑥); and other
predicts values of 𝑥 from given values of 𝑦 (called
line of regression of 𝑥 on 𝑦). Here, Point of
intersection of two lines of regression is (𝑥, 𝑦), Where 𝑥 and 𝑦 are the means of 𝑥 series and
𝑦 series.
Sample Problem: Find the regression line of y on x from the following data:
x 1 3 4 6 8 9 11 14
y 1 2 4 4 5 7 8 9
Solution: Let line of regression of y on x be:
𝑦 = 𝑎 + 𝑏𝑥 … ①
Then normal equations are given by:
∑ 𝑦 = 𝑎𝑛 + 𝑏 ∑ 𝑥 … ② and ∑ 𝑥𝑦 = 𝑎 ∑ 𝑥 + 𝑏 ∑ 𝑥 2 … ③
Calculating∑ 𝑥, ∑ 𝑦, ∑ 𝑥𝑦 and ∑ 𝑥 2
𝑥 𝑦 𝑥2 𝑥𝑦
1 1 1 1
3 2 9 6
4 4 16 16
6 4 36 24
8 5 64 40
9 7 81 63
11 8 121 88
14 9 196 126
∑ 𝑥= 56 ∑ 𝑦= 40 ∑ 𝑥 2 = 524 ∑ 𝑥y = 364
Substituting values of∑ 𝑥, ∑ 𝑦, ∑ 𝑥𝑦 and ∑ 𝑥 2 in ② and③
⇒ 40 = 8𝑎 + 56𝑏 … ④
[27]
 and 364 = 56𝑎 + 524𝑏 … ⑤
6 7
Solving ④and ⑤, we get 𝑎 = and 𝑏 =
11 11
6 7
Substituting in①, line of regression of y on x is𝑦 = 11 + 11 𝑥

Algorithm for fitting line of regression: 𝒚 = 𝒂 + 𝒃𝒙

 Enter the number of data points
 Enter the elements 𝑥𝑖 using loop
 Enter the corresponding points 𝑦𝑖 using loop
 Evaluate ∑ 𝑥, ∑ 𝑦, ∑ 𝑥𝑦 and ∑ 𝑥 2
 Solve the normal equations ∑ 𝑦 = 𝑎𝑛 + 𝑏 ∑ 𝑥 and ∑ 𝑥𝑦 = 𝑎 ∑ 𝑥 + 𝑏 ∑ 𝑥 2 to find the
values of 𝑎 and 𝑏.
 The line of regression of y on x is: 𝑦 = 𝑎 + 𝑏𝑥
 Define the x range and plot the line of regression usingplot (𝑥, 𝑎 + 𝑏 ∗ 𝑥) command
CODE:
clc
n=input('Enter the number of data points:')
printf(' Enter the values of xi')
for i=1:n
x(i)=input('\');
end
printf(' Enter the values of yi')
for i=1:n
y(i)=input('\');
end
sumx=0;sumy=0;sumxy=0;sumx2=0;
for i=1:n
sumx=sumx+x(i);
sumx2=sumx2+x(i)*x(i);
sumy=sumy+y(i);
sumxy=sumxy+x(i)*y(i);
end
a=((sumx2*sumy-sumx*sumxy)*1.0/(n*sumx2-sumx*sumx)*1.0);
b=((n*sumxy-sumx*sumy)*1.0/(n*sumx2-sumx*sumx)*1.0);
[28]
printf('The line is Y=%3.3f +%3.3f X',a,b)
x=(0:2:20)
plot(x, a+b*x)

[29]
[30]
EXPERIMENT 8: PERFORM A CHI-SQUARE TEST FOR GOODNESS TO FIT FOR A GIVEN DATA

Aim/Objective: Program to test goodness of fit for a given dataset using Chi-square test
Course Outcome: CO3
Software Used: SCILAB
Theory: Chi-square (𝜒 2 ) is a right tailed test and describes the magnitude of discrepancy
between theory and observation. If 𝜒 2 = 0; the observed and expected frequencies
completely coincide, ∴ 𝜒 2 provides a measure of correspondence between theory and
observations. It is one of the simplest and most general tests and can be used to perform
1. test of significance of sample variance
2. test of goodness of fit
3. test of independence of attributes in a contingency table
Chi-Square (𝝌𝟐 ) Test for Testing Goodness of Fit
If 𝑂𝑖 (𝑖 = 1,2, … , 𝑛) be a set of observed (experimental) frequencies and 𝐸𝑖 (𝑖 = 1,2, … , 𝑛) be
the corresponding set of expected (theoretical) frequencies,
(𝑂𝑖 −𝐸𝑖 )2
then 𝜒 2 = ∑𝑛𝑖=1 ; with degrees of freedom (𝜈) = 𝑛 − 1.
𝐸𝑖
There are some underlying conditions for applying 𝜒 2 test such as:
1. Sum of frequencies should be large (at least 50)
2. No theoretical cell-frequencies should be small, if small (less than 5) theoretical
frequencies occur, regrouping of two or more cells must be done before
calculating (𝑂𝑖 − 𝐸𝑖 ). Degree of freedom is determined by the number of classes after
regrouping.
Sample Problem: Apply 𝜒 2 test of goodness to fit for the following data:
Observed frequency: 1 5 20 28 42 22 15 5 2
Theoretical Frequency: 1 6 18 25 40 25 18 6 1
Value of 𝜒 2 for 6, 7, 8 degrees of freedom at 5% significance level are 12.592, 14.067 and
15.507 respectively.
Solution: Since first two and the last two cells frequencies are smaller than prescribed,
regrouping the data to 7 cells:
𝑂𝑖 6 20 28 42 22 15 7
𝐸𝑖 : 7 18 25 40 25 18 7
Degrees of freedom (𝜈) = 7 − 1 = 6
(𝑂𝑖 −𝐸𝑖 )2
𝜒 2 = ∑𝑛𝑖=1 𝐸𝑖
(6−7)2 (20−18)2 (15−18)2 (7−7)2
= + + ⋯+ + = 1.685
7 18 18 7
Table value of 𝜒 2 for 6 degrees of freedom at 5% level = 12.592
Calculated value of 𝜒 2 is much less than table value ∴ the fit is very good.

[31]
Algorithm:

1. Input observed frequencies, expected frequencies and degrees of freedom (𝑑. 𝑓. = 𝑛 − 1)

where 𝑛 is the number of categories.

2. Compute Chi-square statistic (𝜒 2 )

3. Compare the calculated value of 𝜒 2 with the critical value from the chi-square table for
the given degree of freedom and significance level α.

4. Decision:

 If 𝜒 2 (𝑐𝑎𝑙𝑐𝑢𝑙𝑎𝑡𝑒𝑑) > 𝜒 2 (𝑐𝑟𝑖𝑡𝑖𝑐𝑎𝑙), reject the null hypothesis 𝐻0 .

 Otherwise, fail to reject 𝐻0 .

5. Output:

 The 𝜒 2 value and whether 𝐻0 is rejected or not.

CODE:

// Chi-Square Test for Goodness of Fit

// Input observed and expected frequencies

observed=[50,30,20]; // Replace with your dataset
expected=[40,40,20]; // Replace with your dataset

// Check if lengths match

if size (observed,"*")<>size(expected,"*")then
error ("The observed and expected frequencies must have the same size.");
end

// Calculate chi-square statistic

chi_square_stat=sum(((observed-expected).^2)./expected);

// Degrees of freedom
n=length(observed);
df=n-1;

// Significance level
alpha=0.05;

// Critical value from chi-square distribution

chi_square_critical=cdfchi("PQ",1-alpha,df);

[32]
// Output results
disp("Chi-Square Statistic: "+string(chi_square_stat));
disp("Degrees of Freedom: "+string(df));
disp("Critical Value: "+string(chi_square_critical));

// Decision
ifchi_square_stat>chi_square_criticalthen
disp("Reject the null hypothesis (H0): The data does not fit the expected distribution.");
else
disp("Fail to reject the null hypothesis (H0): The data fits the expected distribution.");
end

[33]
EXPERIMENT 9: PERFORM T-TEST FOR COMPARING MEANS OF TWO SAMPLES

Aim/Objective: Program to use t-test for comparing means of two samples

Course Outcome: CO3
Software Used: SCILAB
Theory: Gosset; who wrote under the pen-name ‘Student’, derived a theoretical distribution
known as Student’s 𝑡 distribution; which is used to test a hypothesis when the sample size is
small and the population variance is not known. As 𝑛 (size of the sample) increases;
𝑡distribution tends to normal distribution. It is evident from the graph below that t-
distribution has proportionally larger area at its tails than the normal distribution.
If 𝑥1 , 𝑥2 , ⋯, 𝑥𝑛 be a random sample of size 𝑛 ≤ 30, from a normal distribution with mean 𝜇
and variance 𝜎 2 (not known); also 𝑥 be the sample mean and 𝑠 standard deviation of the
sample
𝑥−𝜇 ∑𝑥 ∑(𝑥−𝑥)2
then 𝑡 = 𝑠 , where 𝑥 = , 𝑠=√
𝑛 𝑛−1
√𝑛

Degree of freedom is given by 𝜈 = 𝑛 − 1

Statistics 𝑡 is used to test the significance of:
1. Sample mean
2. Difference between two sample means
3. Sample coefficient of correlation
4. Sample Regression Coefficient

Testing difference between means of two small

samples
Let two independent samples ( 𝑥1 , 𝑥2 , ⋯, 𝑥 𝑛1 ) ; ( 𝑦1 , 𝑦2 , ⋯, 𝑦 𝑛2 ) of sizes 𝑛1 ; 𝑛2 with
∑(𝑥−𝑥)2 ∑(𝑦−𝑦)2
means 𝑥 ; 𝑦 and standard deviations 𝑠𝑥 = √ ; 𝑠𝑦 = √ be drawn from two
𝑛1 −1 𝑛2−1
2
normal populations with means 𝜇1 ; 𝜇2 and having same variance 𝜎 , then to test whether
the two population means 𝜇1 and 𝜇2 are same:
𝑥−𝑦 ∑(𝑥−𝑥)2 +∑(𝑦−𝑦)2 (𝑛1−1)𝑠𝑥2 +(𝑛2−1)𝑠𝑦2
Calculate 𝑡 = 1 1
; where 𝑠 = √ or √
𝑠√ + 𝑛1 +𝑛2 −2 𝑛1 +𝑛2 −2
𝑛1 𝑛2

and compare with table value of 𝑡 (say 𝑡1 ) at (𝑛1 + 𝑛2 − 2) degrees of freedom. Hypothesis
is accepted if calculated value of |𝑡| < 𝑡1 .
Sample Problem: Two independent samples showing weights in ounces of eight and seven
items are given below:
Sample I: 10 12 13 11 15 9 12 14
Sample II: 9 11 10 13 9 8 10
Is the difference between the means of the samples significant?
Value of 𝑡 at 5% level of significance for 15 degrees of freedom is 2.16.

[34]
 Solution: Let the null hypothesis (𝐻0 ) be: 𝜇1 ≈ 𝜇2
Here 𝑛1 = 8 and 𝑛2 = 7, calculating sample means and standard deviations
𝑥 (𝑥 − 𝑥) (𝑥 − 𝑥)2 𝑦 (𝑦 − 𝑦) (𝑦 − 𝑦)2
10 -2 4 9 -1 1
12 0 0 11 1 1
13 1 1 10 0 0
11 -1 1 13 3 9
15 3 9 9 -1 1
9 -3 9 8 -2 4
12 0 0 10 0 0
14 2 4
∑ 𝑥 = 96 ∑ 𝑦 = 70
96 28 70 16
𝑥= = 12 𝑦= = 10
8 7
∑(𝑥−𝑥)2+∑(𝑦−𝑦)2 ∑(𝑥−12)2+∑(𝑦−10)2 28+16
𝑠=√ =√ √8+7−2 = 1.8
𝑛1+𝑛2 −2 8+7−2
𝑥−𝑦 12−10
∴𝑡= 1 1
= 1 1
= 2.15
𝑠√ + (1.8)√ +
𝑛1 𝑛2 8 7

Table value of 𝑡 at 5% level of significance for 13 degrees of freedom is 2.16

∴ Calculated value of |𝑡| < table value of 𝑡
Hence the hypothesis (𝐻0 ) is accepted that 𝜇1 ≈ 𝜇2 , i.e. difference between the means of
the two samples is not significant.
Algorithm:

1. Input Data: Sample 1 and Sample 2.

2. Calculate Sample Means.

3. Calculate Sample Standard Deviations.

4. Calculate the Standard Error SE of the Difference in Means.

5. Calculate the T-Statistic T.

6. Determine Degrees of Freedom df.

7. Compare the T-Statistic with the Critical Value.

 Use the t-distribution table or calculate the p-value for the given degree of freedom
and significance level (e.g., 𝛼 = 0.05)

[35]
  If the absolute value of the T-statistic exceeds the critical value (or p-value is less
than 𝛼), reject the null hypothesis.

CODE:

// T-Test for Comparing the Means of Two Samples

// Input data: Two samples

X1=[12,15,14,10,16];// Sample 1 data
X2=[22,25,20,18,21];// Sample 2 data

// Step 1: Calculate the means of both samples

mean_X1=mean(X1);
mean_X2=mean(X2);

// Step 2: Calculate the standard deviations of both samples

std_X1=stdev(X1);
std_X2=stdev(X2);

// Step 3: Calculate the standard error of the difference in means

n1=length(X1);// Sample size of X1
n2=length(X2);// Sample size of X2
SE=sqrt((std_X1^2/n1)+(std_X2^2/n2));

// Step 4: Calculate the T-statistic

T=(mean_X1-mean_X2)/SE;

// Step 5: Calculate the degrees of freedom

df=((std_X1^2/n1+std_X2^2/n2)^2)/(((std_X1^2/n1)^2)/(n1-1)+((std_X2^2/n2)^2)/(n2-1));

// Step 6: Output the results

disp("T-statistic: "+string(T));
disp("Degrees of Freedom: "+string(df));

// Step 7: Calculate the p-value for the T-statistic (two-tailed test)

p_value=2*(1-cdft("PQ",abs(T),df));// tcdf is the cumulative t-distribution function
disp("P-value: "+string(p_value));

// Step 8: Conclusion based on significance level (e.g., alpha = 0.05)

alpha=0.05;// Significance level
ifp_value<alphathen
disp("Reject the null hypothesis: There is a significant difference between the means.");
else
disp("Fail to reject the null hypothesis: No significant difference between the means.");
end

[36]
[37]
EXPERIMENT 10: SOLVING A LPP USING SIMPLEX METHOD

Aim/Objective: Display First Simplex Table of a Linear Programming Problem

Course Outcome: CO4
Software Used: SCILAB
Theory: Simplex method is used for solving linear programming problems having two or
more variables. It tests adjacent vertices of the feasible set so that at each new vertex, the
value of objective function is improved until an optimal solution is attained.
Sample problem: Solve the linear programming problem:
Max 𝑍 = 3𝑥1 + 2𝑥2
Subject to
𝑥1 + 𝑥2 ≤ 4
𝑥1 − 𝑥2 ≤ 2
𝑥1 , 𝑥2 ≥ 0
Solution: Given lpp is of maximization type and also all the bi’s are positive.
Adding slack variables to constraints, lpp is given by:
Max Z= 3𝑥1 + 2𝑥2 + 0𝑠1 + 0𝑠2
Subject to
𝑥1 + 𝑥2 + 𝑠1 = 4
𝑥1 − 𝑥2 + 𝑠2 = 2
𝑥1 , 𝑥2 , 𝑠1 , 𝑠2 ≥ 0
The simplex table is given as:
Cj → 3 2 0 0
𝑉𝐵
B.V. CBVB 𝑥1 𝑥2 𝑠1 𝑠2 Min Ratio , 𝑥𝑘 > 0
𝑥𝑘

𝑠1 0 4 1 1 1 0 4
𝑠2 0 2 1 -1 0 1 2→
Z=0 -3 -2 0 0 ∆𝑗 = 𝐶𝐵 𝑥𝑗 − 𝐶𝑗
𝑥1 enters and 𝑠2 leaves
𝑠1 0 2 0 2 1 -1 →
𝑥1 3 2 1 -1 0 1
Z=6 0 -5 0 3 ∆𝑗 = 𝐶𝐵 𝑥𝑗 − 𝐶𝑗
𝑥2 enters and 𝑠1 leaves

[38]
 𝑥2 21 0 1 1 1
−
2 2
𝑥1 33 1 0 1 1
2 2
Z = 11 5 1 ∆𝑗 ≥ 0
0 0 22

∴ Maximum Z = 11 at 𝑥1 = 3 and 𝑥2 = 1
Algorithm to Display First Simplex Table of a Linear Programing Problem
 Enter the number of variables in the objective function and the number of constraints
 Using loops enter all the coefficients of the objective function and constraint
equations
 Make the choices for minimization or maximization by entering 1 or 2
 Make the choices for ≤ or ≥ signs by entering 1 or 2
 For each ≤ signs make a column for slack variable
 Print the first simplex table using all the given values

CODE:
clc
clear
p=input("Enter number of varibles in the objective function :");
q=input("Enter number of constraint equations :");
disp('Enter coefficients of the objective function :')
for i=1:p
A(i)=input('\');
end
m=input("Enter 1 to minimize or 2 to maximize");
disp('Enter constraint equations :')
for i=1:q
disp('enter ' + string(i)+ ' equation: ')
for j=1:p
disp('enter coefficient ' + string(j)+ ' :')
B(i,j)=input('\');
if(j==p) then
printf("\n whether you want to maximize or minimize this equation:\nEnter \n 1. for<= \n 2.
for >= :")
C(i)=input('\');
printf("enter constant");
D(i)=input('\');
[39]
 end
end
end

printf("\n\n The LPP is ::::\n\n\n");

if(m==1) then
printf("MIN \n");
else
printf("MAX \n");
end
for i=1:p
if(i==p) then
printf("%dx%d",A(i),i);
else
printf("%dx%d + ",A(i),i);
end
end
printf("\nConstraint equations are \n");
for i=1:q
for j=1:p

printf("%dx%d + ",B(i,j),j);
if(j==p) then
if(C(i)==1) then
printf(" <= %d",D(i));
else if(C(i)) then
printf(" >= %d",D(i));
end
end
end
end
printf("\n");
end
printf("\n\n====================SIMPLEX TABLE
IS=======================\n\n\n");
printf(" CJ |");
for i=1:p-1
printf(" %d",A(i));
end
for i=1:q-1

[40]
printf(" 0s%d",i);
end
printf("\n---------------------------------------------------------");
printf("\nBVCbXb |");
for i=1:p
printf(" x%d",i);
end
for i=1:q
printf(" s%d",i);
end
printf(" Min Xb/x");
printf("\n---------------------------------------------------------\n");
for i=1:p
printf("s%d 0 %d | ",i,D(i));
for j=1:q
printf("%d ",B(i,j));
end
for j=1:q
if(j==i) then
printf("1 ");
else
printf("0 ");
end
end
printf("\n");
end
printf("--------------------------------------------------------");
printf("\n ZJ-CJ |");
for i=1:p
printf(" -%d",A(i));
end
for i=1:q
printf(" 0");
end

[41]
[42]
[43]
EXPERIMENT 11: SOLVING TRANSPORTATION PROBLEM

Aim/Objective: Solve a 3 by 3 transportation problem using MODI method with initial

basic feasible solution by North West Corner Rule
Course Outcome: CO4
Software Used: SCILAB
Theory: Transportation problem is a special case of linear programming which aims to
minimize the transportation cost to supply goods from various sources to different
destinations, while satisfying the supply limit and demand requirement.
In general, if there be m sources and n destinations with ai availability in ith source and
bj requirement in jth destination. Also ‘𝑐𝑖𝑗 ’ is the cost of transportation from ith source to jth
destination, then a transportation problem
seeks to determine non- negative values
of ‘ 𝑥𝑖𝑗 ’, so as to
Minimize Z = ∑𝑚 𝑛
𝑖=1 ∑𝑗=1 𝑥𝑖𝑗 𝑐𝑖𝑗

such that ∑𝑛𝑗=1 𝑥𝑖𝑗 = 𝑎𝑖 fori = 1,….,m

∑𝑚
𝑖=1 𝑥𝑖𝑗 = 𝑏𝑗 forj = 1,….,n

𝑥𝑖𝑗 ≥ 0∀i, j
Also for a balanced transportation
problem∑𝑚 𝑛
𝑖=1 𝑎𝑖 = ∑𝑗=1 𝑏𝑗

Basic feasible solution: A feasible

solution of a 𝑚 × 𝑛 transportation problemin which allocations ‘𝑥𝑖𝑗 ’ are provided satisfying
the conditions ∑𝑛𝑗=1 𝑥𝑖𝑗 = 𝑎𝑖 and ∑𝑚
𝑖=1 𝑥𝑖𝑗 = 𝑏𝑗 for each i and j , is said to be a basic feasible
solution.
Optimal solution: A basic feasible solution which minimizes the total transportation cost is
known as an optimal solution.
Sample Problem1: Obtain an initial basic feasible solution to the following transportation
problem by (i)North West Corner rule (ii)Vogel Approximation method

w1 w2 w3 w4 ai↓
S1 19 30 50 10 7
S2 70 30 40 60 9
S3 40 8 70 20 18
bj  5 8 7 14
Solution: (i) Northwest Corner ruleGiven transportation problem is already
balanced, ∴ allocating min [5,7] =5 to cell (1,1) and adjusting supply and demand across the
row and column.

[44]
 w1 w2 w3 w4 ai↓
S1 19 5 30 50 10 7,2
S2 70× 30 40 60 9
S3 40× 8 70 20 18
bj  5 8 7 14

Moving to right hand cell (1,2), allocating min [2,8] =2 and adjusting supply and demand
across the row and column.
w1 w2 w3 w4 ai↓
S1 19 5 30 2 50× 10× 7,2
S2 70× 30 40 60 9
S3 40× 8 70 20 18
bj  5 8,6 7 14

Moving down to cell (2,2), allocating min [6,9] =6 and adjusting supply and demand across
the row & column.
w1 w2 w3 w4 ai↓
S1 19 5 30 2 50× 10× 7,2
S2 70 × 30 6 40 60 9,3
S3 40× 8× 70 20 18
bj  5 8,6 7 14

Moving to right hand cell (2,3), allocating min [3,7] =3and adjusting supply & demand
across the row and column.
w1 w2 w3 w4 ai↓
S1 19 5 30 2 50× 10× 7,2
S2 70 × 30 6 40 3 60× 9,3
S3 40× 8× 70 20 18
bj  5 8,6 7,4 14
Moving down to cell (3,3), allocating min [4,18] =4 and adjusting supply and demand
across the row & column.
w1 w2 w3 w4 ai↓
S1 19 5 30 2 50× 10× 7,2
S2 70 × 30 6 40 3 60× 9,3
S3 40× 8× 70 4 20 18,14
bj  5 8,6 7,4 14
Allocating the balance supply/demand i.e. ‘14’ in the cell (3,4), the initial basic feasible
solution using north west corner rule is given by

[45]
 w1 w2 w3 w4 ai↓
S1 19 5 30 2 50× 10× 7,2
S2 70 × 30 6 40 3 60× 9,3
S3 40× 8× 70 4 20 14 18,14
bj  5 8,6 7, 4 14

Transportation cost = 19(5) + 30(2) + 30(6) + 40(3) + 70(4) + 20(14) = 1015units

(ii) Vogel's Approximation Method (VAM): Writing the difference of minimum cost and
next minimum cost below each column and on the right of each row. Maximum penalty is 22
against w2, so allocating min[8,18]=8 to minimum cost cell (3,2) in w2 column.
w1 w2 w3 w4 ai↓
S1 19 30× 50 10 7 (9)
S2 70 30 × 40 60 9 (10)
S3 40 8 8 70 20 18 , 10 (12)
bj  5 8 7 14
(21) (22) (10) (10)
Again writing the difference of minimum
cost and next minimum cost, skipping allocated cells and crossed out cells, maximum penalty
is 20 against S2 and S3. Taking S3row arbitrarily and allocating min [10,14]=10 to minimum
cost cell (3,4) in S3, also adjusting corresponding supply and demand

w1 w2 w3 w4 ai↓
S1 19 30 × 50 10 7 (9) (9)
S2 70 30 × 40 60 9 (10) (20)
S3 40× 8 8 70× 20 10 18 , 10 (12) (20)×
bj  5 8 7 14, 4
(21) (22) (10) (10)
(21) ×(10) (10)
Again adjusting corresponding supply and demand
w1 w2 w3 w4 ai↓
S1 19 5 30 × 50 10 7,2 (9) (9) (9)
S2 70 × 30 × 40 60 9 (10) (20) (20)
S3 40× 88 70× 20 10 18 , 10 (12) (20) ×
bj  5 8 7 14 , 4
(21) (22) (10) (10)
(21) × (10) (10)
(51) (10) (50)
Rewriting the difference of minimum cost and next minimum cost, skipping allocated cells
and crossed out cells, maximum penalty is 50 against W4.Allocating min[2,4]=2 to minimum
cost cell (1,4) in S1 , also adjusting corresponding supply and demand
w1 w2 w3 w4 ai↓
S1 19 5 30 × 50 10 2 7,2 (9) (9) (9) (40)
[46]
 S2 70 × 30 × 40 60 9 (10) (20) (20) (20)
S3 40× 88 70× 20 10 18 , 10 (12) (20) ×
bj  5 8 7 14 , 4, 2
(21) (22) (10) (10)
(21) × (10) (10)
(51) (10) (50)
(10) (50) (50)

Now allocating remaining demands 7 and 2 of W3 and W4, supply in S2 is exhausted.

w1 w2 w3 w4 ai↓
S119 5 30 × 50× 10 2 7,2 (9) (9) (9) (40)
S2 70× 30 × 40 7 60 2 9 (10) (20) (20) (20)
S340× 8 8 70× 20 10 18 , 10 (12) (20) ×
bj  5 8 7 14 , 4 , 2
(21) (22) (10) (10)
(21) × (10) (10)
(51) (10) (50)
× (10) (50)
Transportation cost = 19(5) + 10(2) + 40(7) + 60(2) + 8(8) + 20(10) = 779units
Sample Problem2: Find an optimal solution of the transportation problem given in Sample
Problem1.
Solution: An initial basic feasible solution of the
transportation w1 w2 w3 w4 problem as given by VAM is:
S1 19 5 30 50 10 𝟐
S2 70 30 40 7 60 𝟐
S3 40 8 𝟖 70 20 𝟏𝟎
To assign set of numbers ui and vj, putting u2 =0 arbitrarily as all the rows are having equal
number of allocations. Calculating remaining ui and vjusing the relation
cij = (ui + vj)for occupied w1 w2 w3 w4 cells.
S1 19∗ 10 ∗ u1 =-50
S2 40 ∗ 60 ∗ u2 = 0
S3 8 ∗ 20 ∗ u3= -40
v1=69 v2=48 v3=40 v4=6
Now entering cijfor each unoccupied cell (i,j)in
the upper left corner and (ui + vj) for the same cell in the upper right corner. Also writing
dij= cij – (ui + vj) at lower right corner. We observe that d22 = -18, seeking an allocation
which is fulfilled by forming a closed loop with other occupied cells travelling horizontally
or vertically.
w1 w2 w3 w4
[47]
 S1 ∗ 30
-2 50 -10 ∗ u1= -50
+32 +60
S2 70 69 30 ? 48 ∗ ∗ u2 = 0
+1 -18
S3 40 29 ∗ 70 0 ∗ u3 = -40
+11 +70
v1=69 v2=48 v3=40 v4=60

Adding and subtracting min[2,8]=2 at vertices of the loop to adjust supply and demand
requirements as given below
?+2 2-2

8-2 10+2
Modified distribution of allocations is as shown in the table below:
w1 w2 w3 w4
S1 19 5 30 50 10 𝟐
S2 70 30 2 40 7 60
S3 40 8𝟔 70 20 𝟏𝟐
New transportation cost = 19(5) + 10(2) + 30(2)+ 40(7) + 8(6)+20(12) = 743units
To check whether the new solution is optimal, assign set of numbers u i and vj, putting u1 =0
arbitrarily as all the rows are having equal number of allocations. Calculating remaining ui
and vjusing the relation cij = (ui + vj)for occupied cells.
w1 w2 w3 w4
S1 19 ∗ 10 ∗ u1 =0
S2 30 ∗ 40 ∗ u2 =32
S3 8 ∗ 20 ∗ u3=10
v1=19 v2=-2 v3=8 v4=10

Now entering cijfor each unoccupied cell (i,j)in the upper left corner and (ui + vj) for the
same cell in the upper right corner.

w1 w2 w3 w4
S1 ∗ 30 -2 50 8 ∗ u1 = 0
+32 +42
S2 70 41 ∗ ∗ 60 42 u2=32
+29 +18
S3 40 29 ∗ 70 18 ∗ u3=10
+11 +52
v1=19 v2=-2 v3=8v4=10
[48]
All dij> 0, ∴ the solution is optimal and minimum transportation cost = 743 units

Algorithm to solve a balanced transportation problem using MODI method withinitial

basic feasible solution by Northwest Corner Rule.

The Procedure is conducted in two steps:

I. Finding an initial basic feasible solution using North West Corner Rule:
1. Balance the transportation problem if not originally by adding a dummy source or
destination making ∑𝑚 𝑛
𝑖=1 𝑎𝑖 = ∑𝑗=1 𝑏𝑗 ,with zero transportation cost in added cells.
2. Start with the cell in the upper left hand corner which is north west corner (1,1)
and allocate the maximum possible amount 𝑥𝑖𝑗 = Min (𝑎𝑖 , 𝑏𝑗 ) in the cell (i, j),
such that either the availability of the source 𝑆𝑖 is exhausted or the requirement at
destination 𝐷𝑗 is satisfied or both.
3. Adjust supply and demand across the row and column in which allocation 𝑥𝑖𝑗 has
been made.

4. Move to right hand cell (1,2) if there is still any available quantity left otherwise
move down to cell (2,1).
5. Continue the procedure until all the available quantity is exhausted.

II. Finding the optimal solution using Modi Method

1. Find an initial basic feasible solution using Northwest Corner Rule
2. Assign a set of numbers, ui(i= 1,2,…,m) with each row and vj(j=1,2,…..,n) with each
column, such that for each occupied cell (r,s), cr,s= ur + vs . Start with a ui =0 for the
row having maximum number of occupied cells and calculate remaining u i and vj
using given relation for only occupied cells.
3. Enter cijfor each unoccupied cell (i,j)in the upper left corner and enter ui + vj for the
same cell in the upper right corner.
4. Enter dij= cij – (ui + vj)for each unoccupied cell at lower right corner.
5. Examine dijfor unoccupied cells.
(i) If all dij> 0, solution is optimal and unique.
(ii) If all dij ≥ 0, solution is optimal and an alternate optimal solution exists.
(iii) If least one dij< 0, solution is not optimal and go to step 6.
6. Form a new basic feasible solution by giving an allocation to the cell for which dijis
most negative.For this unoccupied cell, generate a path forming a closed loop with
occupied cells by drawing horizontal and vertical lines between them. Alternately put
‘+’ and ‘-’ signsat vertices of the loop starting with the unoccupied cell which is
seeking an allocation. Add and subtract the minimum allocation where a subtraction
is to be made so that one of the occupied cells is vacated in this process.
7. Repeat steps 2 to 5 until and optimal solution is obtained.

[49]
CODE:
clc
prices = [1, 2, 3;
8, 5, 4;
3, 1, 6]
demand = [100, 30, 70]
supply = [110, 40, 50]
prices = evstr(x_matrix('setprices', prices));
demand = evstr(x_matrix('setdemand', demand));
supply = evstr(x_matrix('submit offer', supply));
LEFT = 1
RIGHT = 2
UP = 3
DOWN = 4
// The function of calculating the cost of the transferred plan
function res=cost(prices, plan)
cntCols = length(prices(1,:))
cntRows = length(prices(:,1))
res = 0
for i=1:cntRows
for j=1:cntCols
res = res + prices(i,j) * plan(i,j)
end
end
endfunction
// a function that looks for available angles in a given direction
// and returns them in descending order of proximity to the edge
function [corners, success]=getAvailableCorner(basis, direction, initialPoint, i, j)
success = 0
corners = []
currentCorner = 1
cntCols = length(basis(1,:))
cntRows = length(basis(:,1))
colModificator = 0;
[50]
rowModificator = 0;
if direction == LEFT then
colModificator = -1
end
if direction == RIGHT then
colModificator = 1
end
if direction == UP then
rowModificator = -1
end
if direction == DOWN then
rowModificator = 1
end
i = i + rowModificator
j = j + colModificator
while i ~= 0 && j ~= 0 &&i<= cntRows&& j <= cntCols
if basis(i,j) ~= 0 || [i, j] == initialPoint then
corners(currentCorner,:) = [i,j]
currentCorner = currentCorner + 1
success = 1
end
i = i + rowModificator
j = j + colModificator
end

if success == 1 then
cornersReverse = []
for iter = 1:length(corners(:,1))
cornersReverse(iter,:) = corners(length(corners(:,1)) - iter + 1, :)
end
corners = cornersReverse
end
endfunction
// recursive looping function
[51]
function [nodes, success]=buildCycle(basis, initialPoint, currentPoint, direction)
success = 0
nodes = []
possibleDirections = []
if initialPoint == currentPoint then
possibleDirections = [LEFT, RIGHT, UP, DOWN]
else if direction == LEFT || direction == RIGHT then
possibleDirections = [UP, DOWN]
else if direction == UP || direction == DOWN then
possibleDirections = [LEFT, RIGHT]
end; end; end
for directionIdx = 1:length(possibleDirections)
[corners, suc] = getAvailableCorner(basis, possibleDirections(directionIdx), initialPoint,
currentPoint(1), currentPoint(2))
if suc == 1 then
possibleToCloseCycle = 0
successWithCorners = 0
for cornIdx = 1:length(corners(:,1))
if (corners(cornIdx,:) == initialPoint) then
possibleToCloseCycle = 1
continue
end
[subNodes, suc] = buildCycle(basis, initialPoint, corners(cornIdx,:),
possibleDirections(directionIdx))
if suc == 1 then
successWithCorners = 1
nodeIdx = 1
nodes(nodeIdx, :) = currentPoint
for subNodeIdx = 1:length(subNodes(:,1))
nodeIdx = nodeIdx + 1
nodes(nodeIdx, :) = subNodes(subNodeIdx,:)
end
break
end

[52]
 end
if successWithCorners == 1 then
success = 1
break
else if possibleToCloseCycle == 1 then
nodes(1, :) = currentPoint
nodes(2, :) = initialPoint
success = 1
break
end; end
end
end
endfunction
cntCols = length(prices(1,:))
cntRows = length(prices(:,1))
plan = [] // reference plan
plan(cntRows, cntCols) = 0 // fill with zeros
// Calculation of the original reference plan using the northwest corner method
tempDemand = demand
tempSupply = supply
for j=1:cntCols// iterate over columns (customers)
for i=1:cntRows// iterate over rows (suppliers)
currentSupply = min(tempDemand(j), tempSupply(i))
plan(i,j) = currentSupply
tempDemand(j) = tempDemand(j) - currentSupply
tempSupply(i) = tempSupply(i) - currentSupply

if tempDemand(j) == 0 then
break
end
end
end
disp("Initial plan:")
disp(plan)
[53]
printf("\nThe cost is %d у.е.\n\n\n", cost(prices, plan))
// Plan optimization
optimal = 0
UNKNOWN_POTENCIAL = 9999999
iteration = 0
while optimal ~= 1
iteration = iteration + 1
potencialU = []
potencialV = []
for i = 1:cntRows
potencialU(i) = UNKNOWN_POTENCIAL // type unknown yet potential
end
for i = 1:cntCols
potencialV(i) = UNKNOWN_POTENCIAL
end
potencialU(1) = 0
continuePotentialing = 1
// calculation of potentials by points in the route
while continuePotentialing == 1
continuePotentialing = 0
// we continue to calculate the potentials if
// for one of the values ​​of the plan, both potentials are unknown
for j=1:cntCols// iterate over columns (customers)
for i=1:cntRows// iterate over rows (suppliers)
if (plan(i,j) == 0) then
continue
end
if potencialU(i) == UNKNOWN_POTENCIAL &&potencialV(j) ==
UNKNOWN_POTENCIAL then
continuePotentialing = 1
continue
end
if potencialU(i) == UNKNOWN_POTENCIAL then
potencialU(i) = prices(i,j) - potencialV(j)

[54]
 end
if potencialV(j) == UNKNOWN_POTENCIAL then
potencialV(j) = prices(i,j) - potencialU(i)
end
end
end
end
// Calculating estimates for non-basic variables
notBasis = [] // reference plan
notBasis(cntRows, cntCols) = 0 // fill with zeros
optimal = 1
maxI = 0;
maxJ = 0;
maxNB = 0;
for j=1:cntCols// iterate over columns (customers)
for i=1:cntRows// iterate over rows (suppliers)
if (plan(i,j) ~= 0) then
continue
end
notBasis(i,j) = potencialU(i) + potencialV(j) - prices(i,j)
if notBasis(i,j) > 0 then
optimal = 0
if maxNB<notBasis(i,j) then
maxNB = notBasis(i,j)
maxI = i
maxJ = j
end
end
end
end
if optimal == 1 then
printf("Iteration %d. The current plan is optimal!", iteration)
break
else
[55]
printf("Iteration %d. Current plan is not optimal. Plan optimization", iteration)
end
[nodes, success] = buildCycle(plan, [maxI, maxJ], [maxI, maxJ], "")
if success == 0 then
disp("Loop building error. Shutdown")
break
end
// Among the even nodes of the cycle (those who will have a negative 0) looking for the
minimum value
minNode = 99999999
for node = 2:2:length(nodes(:,1))
if minNode>plan(nodes(node, 1), nodes(node, 2)) then
minNode = plan(nodes(node, 1), nodes(node, 2))
end
end
for node = 2:length(nodes(:,1))
nodeI = nodes(node, 1)
nodeJ = nodes(node, 2)
if modulo(node, 2) == 0 then
plan(nodeI, nodeJ) = plan(nodeI, nodeJ) - minNode// for even subtract min. meaning
else
plan(nodeI, nodeJ) = plan(nodeI, nodeJ) + minNode// for odd ones add min. meaning
end
end

disp("New plan:")
disp(plan)
printf("\nThe cost is %d у.е.\n\n\n", cost(prices, plan))
end
tableStr = 2;
table = []
table(1,:) = [" " "From supplier" "To the consumer" "Quantity"];
for i = 1:cntRows
for j = 1:cntCols

[56]
 if plan(i,j) ~= 0 then
str = []
str(1) = " "
str(2:4) = string([i, j, plan(i,j)])
table(tableStr,:) = str
tableStr = tableStr + 1
end
end
end
disp(table)
f = createWindow();
f.figure_size = [400 400];
f.figure_name = "Final answer";
as = f.axes_size;
ut = uicontrol("style", "table",..
"string", table,..
"position", [0 -50 400 400],..
"tag", "Final answer");
matrix(ut.string, size(table))

Output:

[57]
[58]
EXPERIMENT 12: SOLVING A 3 BY 3ASSIGNMENT PROBLEM

Aim/Objective: Solving a 3 by 3 assignment problem (Single step assignments)

Course Outcome: CO4
Software Used: SCILAB
Theory: An assignment problem is a particular case of transportation problem in which a
number of operations are to be assigned to an equal number of operators, where each operator
performs only one operation. The objective is to minimize overall cost or to maximize the
overall profit for a given assignment schedule.
Mathematical Representation of an Assignment Problem
If there are n jobs to be performed and n persons are available for doing this job. Assume all
persons can do each job at a time with different payoffs. Let c ij be the cost associated with ith
person assigned with jth job. Then the problem is to find the values ‘ 𝑥𝑖𝑗 ’ (𝑥𝑖𝑗 = 0 or 1), so
that total cost for performing all the jobs is minimum.
Minimize Z = ∑𝑚 𝑛
𝑖=1 ∑𝑗=1 𝑐𝑖𝑗 𝑥𝑖𝑗

Such that
∑𝑛𝑗=1 𝑥𝑖𝑗 = 1 for i = 1,….,n
∑𝑛𝑖=1 𝑥𝑖𝑗 = 1for j = 1,….,n
𝑥𝑖𝑗 = 1if ith person is assigned with jth job, 0 otherwise
Algorithm for solving a 3 by 3 assignment problem (Single step assignments)
1. Prepare a square 3 × 3 matrix. If not, make it square by adding suitable number of
dummy rows (or columns) with zero cost elements.
2. Subtract the minimum element of each row from every element of that row.
3. Further modify the resulting matrix by subtracting the minimum element of each
column from all the elements of the respective columns, so that each row and column
is having a zero element.
4. Now start making assignments row - wise. Examine each row one by one and assign a
‘1’’ to ‘0’ in the rows having single zeros. Then, mark a ‘×’ to all zeroes in the
column in which assignment has been made so that no other assignment can be made
in the same column
5. Repeat the procedure for the columns.
6. Stop if all 3assignments have been made.

[59]
 Sample problem: A department head has four subordinates and four tasks have to be
performed. Time (hours) each man would take to perform each task is given below. How
the tasks should be allocated to each subordinate so as to minimize the total man-hours?
Subordinates
S1 S2 S3 S4
Tasks T1 5 6 8 9
T2 6 8 10 6
T3 9 5 8 5
T4 9 8 7 10

Solution: Given assignment problem is balanced i.e. number of rows and columns are equal.
Subtracting minimum element in each row from all the elements of the row, hence inducing a
zero element in each row
S1 S2 S3 S4
T1 0 1 3 4
T2 0 2 4 0
T3 4 0 3 0
T4 2 1 0 3
Each column is having a zero element, so matrix is unchanged by subtracting minimum
element in each column. Now making assignments on zeros to rows having single zeros and
crossing out remaining zeros in same columns and repeating the process with columns if any
allocations are left in row assignments
S1 S2 S3 S4
T1 0 1 3 4
T2 2 4 0
T3 4 0 3
T4 2 1 0 3
All the assignments have been made in single step.
Optimal assignments: T1 - S1 , T2 - S4 , T3 - S2 , T4 - S3
Minimum man hours = (5+6+5+7) hours = 23 hours

[60]
CODE:

clc
clear
n=input("how many workers and job do you have :");
printf("\nenter time of\n\n");
for i=1:n
for j=1:n
printf("worker %d job%d :",i,j);
T(i,j)=input('\');
end
end
printf("\nDATA YOU ENTERED IS :\n ");
for i=1:n
printf("job%d ",i);
end
for i=1:n
printf("\nworker%d ",i);
for j=1:n
printf("%d ",T(i,j));
end
end
minim=[1000,1000,1000,1000,1000,1000,1000,1000,1000];
for i=1:n
for j=1:n
if(T(i,j)<=minim(i)) then
minim(i)=T(i,j);
end
end
end
printf("\n");
for i=1:n
for j=1:n
T(i,j)=T(i,j)-minim(i);
end

[61]
end
printf("\n");
printf("\n\n*************data after row minimum decrement is***************\n\t\t");
for i=1:n
printf("job%d ",i);
end
for i=1:n
printf("\nworker%d ",i);
for j=1:n
printf("%d ",T(i,j));
end
end
zerr=[1000,1000,1000,1000,1000,1000];
zerc=[1000,1000,1000,1000,1000,1000];
for i=1:n
for j=1:n
if(T(i,j)==0) then
zerr(i)=0;
zerc(j)=0;
end
end
end
f=0;
y=0;
for i=1:n
if(zerr(i)) then
f=1;
end
end
for i=1:n
if(zerc(i)) then
y=1;
end
end
if((f==1)||(y==1)) then

[62]
mn=[1000,1000,1000,1000,1000,1000,1000,1000,1000];

for j=1:n
for i=1:n
if(T(i,j)<=mn(j)) then
mn(j)=T(i,j);
end
end
end
for j=1:n
for i=1:n
T(i,j)=T(i,j)-mn(j);
end
end
printf("\n\n*************data after column minimum decrement is***************\n");
for i=1:n
printf("job%d ",i);
end
for i=1:n
printf("\nworker%d ",i);
for j=1:n
printf("%d ",T(i,j));
end
end
end
printf("\n\n\n************Final job assignment is**************");
for i=1:n
for j=1:n
if(T(i,j)==0) then
printf("\n assign job %d to worker %d",j,i);
for z=1:n
if (z==i) then
continue;
end
if(T(z,j)==0) then

[63]
 T(z,j)=1000;
end
for l=1:n
if(l==j) then
continue;
end
if(T(i,l)==0) then
T(i,l)=1000;
end
end
end
end
end
end
printf(“\n");

[64]
[65]